Postgraduate Session at ICTD2010: ICT projects in rural areas of developing countries have an extremely low success ratio in terms of sustainability, especially in the field of Public Health. How to achieve sustainability in ongoing ICT interventions and prevent failure is an open issue. This work suggests a methodology for needs assessment regarding sustainability and for the design of a Sustainability Action Plan to be applied to EHAS-Napo project, an e-Health initiative in the Peruvian Amazon. The Sustainable ICT Framework has been applied in order to improve the project sustainability expectations in the medium term.

1. Postgraduate strand at ICTD2010 1
Design of a Sustainability Action Plan for
EHAS-Napo project: a rural e-Health initiative
Inés Bebea González Leopoldo Liñán Benítez Carlos Rey Moreno
Abstract—ICT projects in rural areas of developing countries
have an extremely low success ratio in terms of sustainability,
especially in the field of Public Health. How to achieve
sustainability in ongoing ICT interventions and prevent failure is
an open issue. This work suggests a methodology for needs
assessment regarding sustainability and for the design of a
Sustainability Action Plan to be applied to EHAS-Napo project,
an e-Health initiative in the Peruvian Amazon. The Sustainable
ICT Framework has been applied in order to improve the project
sustainability expectations in the medium term.
Index Terms—ICT, e-Health, sustainability, maintenance,
developing countries, rural
Fig. 1. Napo project in Peruvian Amazon communicates rural health
facilities with city hospital.
I. INTRODUCTION
R ural areas of developing countries are the living
context of 3 billion people. Access to the information
society is extremely difficult in such environments, contribute to the overall objective of the development
characterized by a lack of infrastructure for communications as intervention because the project is abandoned in the medium
well as for roads and electricity, scarcity of qualified staff able term.
to handle technology and scattered low-income population.
This makes it difficult to both launch and maintain Information By this definition, an ICT project is sustainable if it
and Communication Technologies (ICTs) over time, which maintains and extends its benefits over the medium to long
leads to sustainability. term [7]. The main difficulty in achieving this sustainability
arises from the fact that sustainability needs to be
Despite ICTs being proposed as a cross-cutting and multi- simultaneously achieved in several categories: economical,
sectoral approach to promote social priorities for achieving the financial, institutional, technological, educational, and also
Millennium Development Goals (MDGs) and millions of US- social and cultural challenges [8]. In the particular case of
dollars being invested in this area, initiatives in developing ICTs applied to Public Health, sustainability additionally
countries have shown extremely low success ratios in terms of encounters the challenge to support and facilitate
sustainability. Some studies claim high failure rates [1-5] up to organizational change of Health Institutions [9], [10]. Failure
80% [6]. After literature review of success and failure of is also prevalent in this effort [11], [12].
development ICT interventions, we found difficulties related
to the lack of reported project evaluation, the lack of Fundación EHAS is an international NGO promoting
specialized literature where case studies abound, and disparate telemedicine un rural isolated areas in developing countries,
stakeholders interests and evaluation moments. Following being the Amazon region a well-known environment [13],
Heeks [6] definitions, we consider sustainability failure as a [14], [15]. EHAS-Napo project seems to be a successful
successful initiative that achieves its expected results and initiative while the NGO is still working on the field.
therefore manages to reach its explicit objective, but fails to However, some factors have been identified that indicate that
the mid-term risk of sustainability failure is high. In this work,
I analyze and propose a solution for the sustainability problem
Manuscript received July 31, 2010. of EHAS-Napo project. This initiative started in 2007 and
Inés Bebea González and Carlos Rey Moreno are researchers at serves a broadband wireless network for the National Health
Fundacion EHAS and MsC. in Telecommunication Networks for Developing
Countries by Universidad Rey Juan Carlos in 2010 (e-mail: System in the rural Amazonic region of Loreto, Peru. This
ines.bebea@ehas.org, crey@ehas.org). project brings autonomous communication to 18 isolated
Leopoldo Liñán Benítez is a researcher at Rural Telecommunications health facilities along Napo river (covering a distance larger
Group in Pontificia Universidad Católica in Peru (linan.el@pucp.edu.pe).

2. Postgraduate strand at ICTD2010 2
than 500 km) from Iquitos Regional Hospital to Cabo Pantoja therefore to 60% in data and telemedicine services.
in the frontier to Ecuador (see Figure 1). Connectivity is Maintenance difficulties also cited by Surana [19]
achieved using WiLD (WiFi IEEE 802.11 for Long Distance) lead to high MTTR (Mean Time to Recovery) due to
technologies, and also some services are provided on top of extremely high cost of river transport, insufficient
the network, such as VoIP telephony, videoconferencing, budget for spare equipment and dependence on
reporting, chat and Internet access among others. experts for fault diagnosis.
Financial.- Refers to economic mechanisms to
In 2009, Napo project seemed prepared for transfer process support initial investment, operational (software,
to public institutions, as the initiative counted with strong licenses, management) and maintenance costs
support and motivation from rural health staff and (replacing equipment included) taking into account
municipalities. However, low availability of e-Health services Total Cost of Ownership [20]. The main financial
and local technicians dependence on NGO engineers, pointed problem identified in EHAS-Napo project is that no
to a maintenance weakness. However, a broader analysis permanent budget for ICT operation and maintenance
conducted to study the problem of sustainability as a whole. guaranteed by public institutions.
Social.- Includes social and institutional arrangements
The aim of this work is both to identify risks related to to preserve ICT use for its intended social benefits.
sustainability and relevant criteria in the design of an action Project legal and political framework includes
plan that improves the present situation with a view to telecom market and radio-electric spectrum
definitive ICT project delivery. regulations and information policies that the project
must assess in order to be independent of future
political decisions or institutional changes. Social
II. METHODOLOGY environment includes local mobilization, identifying
To identify and analyze problems related to maintenance needs, ownership, participation and fulfillment of
and all related activities, I conducted a case study qualitative explicit objectives. In the particular case of e-Health
research methodology that included the following: a revision projects, this entails commitment of health staff to
of maintenance documentation materials (technical and audit institutionalization and organizational change, which
reports, manuals), statistics of IT services availability and use; is also crucial in Napo project. There, the formal
participant observation in Napo network; in-depth interviews political and legal framework was not fully stable as a
to engineers, doctors responsible for rural health networks, consequence of the ongoing decentralization process
doctors at reference hospital, local and regional authorities of the National Health System.
assistants; as well as inquiries to rural health and maintenance Human.- Refers to individuals knowledge and
staff [16]. previous personal experiences, learning skills and
cultural endowment. In Napo project, use of ICTs is
For the analytic study of compiled information, detected estimated below its potential, due to high staff
factors have been classified following the Sustainable ICT rotation and deficiencies in the limited learning and
Framework [17]. Logical Framework Approach (LFA) has education skills cited above. Identified difficulties in
been used to analyze trees of partial problems and solutions in capacity building include low frequency in-person
the design of the Sustainability Action Plan (SAP): objectives, courses given to users and technicians, courses
expected results, activities and stakeholders involved. SAP is focusing in computing and office applications,
detailed in Master Thesis [16] but only briefly presented here. deficient attendance to courses because of
The plan is design with 3 subplans facing different categories incompatibilities with other functions (i.e. leaving a
of sustainability and focus stakeholders. rural facility unattended) and high costs of in-person
courses due to high transport and expenses costs.
Content.- Refers to the information communicated
III. ANALYSIS OF SUSTAINABILITY FACTORS IN NAPO PROJECT through ICTs [18]: externally produced (i.e. Internet),
Considering the five categories of sustainability [18], [17], externally or locally adapted, and locally produced
the following is a brief explanation of problems detected: (i.e. epidemiological surveillance reports mailing
Technological.- Involves electric, ICT and road from rural health facilities). Content is useful once it
infrastructure. Technology must be robust, low- fulfills language requirements, is understandable in
maintenance and low-cost, while maintenance is the specific literacy and cultural context and answers
highly dependent on equipment availability and road real user needs. This involves also project software
infrastructure. Although ICTs based on WiLD and documentation. In the project, the Health
technologies and autonomous solar electrification Information System has been partially adapted at the
provided by Napo project are robust, availability lower statements, where some information has been
statistics dropped to 70% in connectivity and generated locally. However, production of locally

3. Postgraduate strand at ICTD2010 3
Fig. 2. Sustainability Action Plan improving sustainability at each category.
relevant information in the form of on-line seminars
Fig. 3. Operational Maintenance Executive Core.
or multicast discussions on clinic cases, is still a
challenge in Napo project.
IV. DESIGN OF A SUSTAINABILITY ACTION PLAN maintenance response and recovery time. A Ticketing
SAP is conceived with the aim of improving the System is proposed for this purpose: Request Tracker
sustainability of Napo project. For clearer objective-activity [23], due to its easy connection for listening NMS
design, three subplans are propoposed (see Figure 2): an alerts.
Operative Maintenance Plan (OMP) fighting technological Predictive Maintenance studies statistical behavior
factors focusing in ICT professionals; an Institutional and of relevant parameters over time in order to anticipate
Finantial Plan (IFP) facing social-intitutional difficulties also faults and keep technological surveillance and
related to financing in the case of Public Health Institutions; adaptation through R&D.
and a Continuous Learning Plan (CLP) improving human and Preventive Maintenance plans 6-month mainte-
content-related factors affecting sustainability with a focus nance interventions for general ICT infrastructure,
in training resources. hardware and software. The aim of these actions is to
correct minor faults, prevent service interruptions and
A. Operational Maintenance Plan avoid costly corrective actions.
Corrective Maintenance consists of ad hoc remote
The main goal for the Operative Maintenance Plan is to
or in-person reparations once faults occur. Major
reduce MTTR, thus improving ICT availability to 90% and
costs here come from non-programmed transportation
80% for connectivity and telemedicine services respectively.
to remote nodes.
Proposed protocol for maintenance is shown in Figure 3, with
Stock Management includes control of spare
a central core which is the ’Historical Knowledge of Network
equipment and toolbox, that must be ready for use in
Status’ (HKNS). This core is the basis for accurate, registered
maintenance. Stock input-output recording, and
and up-to-date information that allows fast response,
purchase and shipment of equipment is also ticketed
experience-based fault diagnosis and technology adaptation to
within Incident Management.
dynamic user needs. The protocol consists of the following
items: Reports of all cited actions contribute to HKNS
archive.
Monitoring collects accurate information from
network devices and services running. This R&D allows long-term surveillance on technology
information includes connectivity status, signal and user changing needs over time.
strength received at wireless routers, network traffic,
delay and data rate, CPU consumption, via a Network This subplan also specifies four maintenance levels for
Management System (NMS). Contextual information technical staff according to their (increasing) ICT education
is collected from user polling, thus checking non- and expertise and to their (decreasing) proximity to network
managed service performance and human perception. nodes. Layers are listed below:
The system proposed here is open-source Centreon- L1: half-time local technicians (non ICT professionals, staff
Nagios [21],[22]. at rural health facilities)
L2: full-time operative engineering (ICT professionals,
Incident Management records anomalous events
regional institutions staff at Telecom and Telemedicine
and failures occurring during ICTs operation,
Offices)
including detection (user case or automated alerts
L3: ad-hoc support engineering (ICT professionals, NGO
generated by NMS), diagnosis decisions and
staff from remote)
maintenance actions taken -all items crucial in

4. Postgraduate strand at ICTD2010 4
L4: asynchronous R&D (ICT researches, NGO staff from
remote)
Users are considered at the top of the pyramid, so their
capacity to understand and describe failures is crucial in this
process (see Continuous Learning Plan). Annual maintenance
for Napo project costs 7% of the initial investment in project
implementation.
B. Continuous Learning Plan
Taking into account particularities of rural health staff[24],
CLP defines a learning methodology based in andragogic
model and competence building [25], in-person and (open-
source) distance learning modules, and training of trainers at
Training Unit at the regional health division. Proposed
contents are divided into four main modules. Each module
may count with in-person and distance parts.
M1.- Computing, following International Computer Driving
License (ICDL) [26], for users and technicians (2 levels).
M2.- Use and administration of distance-learning tool for
learners and trainers (2 levels).
M3.- ICT and Telemedicine operation.
M4.- ICT maintenance for users and technicians (2 levels).
Resources needed to implement this subplan are estimated
budget 3% of initial Napo project investment per year.
Restrictions taken in this calculation include: learning tools
(didactic examples, teaching guides and support material), Fig. 4. Process for Institutional Change Management in Napo project.
infrastructure (Moodle server installation and maintenance,
edition and publication costs, computer rooms, projector and
blackboard rentals), and human resources (one trainer each 10
learners and one trainer each 40 online learners).
Institutions at a regional level can be huge, and
organizational changes need presidential approval and 2-year
C. Institutional and Financial Plan
procedure. Also necessary functions for implementing SAP are
Local authorities in Napo project consist of municipalities non-existent in the organigrams of regional institutions, so new
of three districts involved and two rural healthcare networks. offices are proposed to fill this gap. Figure 4 shows the
There also exist major authorities including regional strategy proposed to boost institutional change in the ICT
government and regional health division, which have the project.
maximum administrative and health competencies in the
decentralized national system. These regional institutional V. DISCUSSION ON RESULTS AND FUTURE WORK
bodies need to provide funds for the initiative in the long term.
Then the principal objective in this Plan is to manage Results achieved are hard to extrapolate even to similar
institutional change from three fronts: initiatives due to the idiosyncrasy of e-health projects.
However, the methodology proposed for this case could be
Redistribute responsibilities and roles related to ICT
adapted to address other e-Health and ICT projects: a deep
project by passing them from the NGO to
assessments of factors characterizing sustainability and a
competentregional institutions.
global vision of a plan (much detailed but not included here) to
Regularize formal legal, political and administrative
improve sustainability assessment at any time of ICT project’s
framework of the project, which should be adapted to
life cycle.
the new decentralized model.
Build dispositions to guarantee a permanent budget in
During 2010 SAP has been partially implemented up to
order to meet operational and maintenance costs
25% of programmed activities. Some results have been
(SAP costs are known now as annual 10% of
collected that show partial solutions achieved and that we are
investment) under a regional-local self-management
reducing factors threatening the sustainability of this initiative.
proposal.

5. Postgraduate strand at ICTD2010 5
For a precise evaluation of SAP impact on sustainability, [25] D. Fidishun, “Andragogy and technology: Integrating adult learning theory
as we teach with technology,” Proceedings of the 2000 Mid- South
current work is focused on the design of a Validation Instructional Technology Conference, 2000.
Framework consisting of qualitative and quantitative indicators [26] “International computer driving license syllabus version 5.0,”
http://www.ecdl.com, 2007.
at each category in order to compare previous and later
situations.
REFERENCES
[1] C. Avgerou, “Information systems in developing countries: a critical
research review,” Journal of Information Technology, vol. 23, pp. 133– 146,
Sep. 2008.
[2] K. L. Kraemer, J. Dedrick, and P. Sharma, “One laptop per child: vision vs.
reality,” Commun. ACM, vol. 52, no. 6, pp. 66–73, 2009.
[3] R. Rajagoplan and R. Sarkar, “Digital networks and sustainability: Do we
need the government?” International Conference on the Digital Society, vol.
0, pp. 49–54, 2008.
[4] A. Chib, “Sustainability of ict interventions: Lessons from rural projects in
china and india,” Communicating for Social Impact: Engaging Com-
munication Theory, Research, and Pedagogy, 2009.
[5] F. Salem, “Failed revolution? exploring e-government barriers in the arab
states,” Proceedings of the 4th International Conferrence on e- Government,
pp. 363–370, 2008.
[6] R. B. Heeks, “Information systems and developing countries: Failure,
success and local improvisations,” Inf. Soc., vol. 18, no. 2, 2002.
[7] Batchelor, S. and Norrish, P., “Framework for the assessment of ICT pilot
projects: Beyond monitoring and evaluation to applied research,”infoDev,
World Bank Information and Development Program, 2009. [Online].
Available: http://www.infodev.org/en/Publication.4.htm
[8] C. I. Pade, “An exploration of the categories associated with ict project
sustainability in rural areas of developing countries: a case study of the
dwesa project,” Proceedings of SAICSIT, pp. 100–106, 2006. E. L.
[9] Mosse, “Understanding the introduction of computer-based health
information systems in developing countries: counter networks, commu-
nication practices and social identity. a case study from mozambique,” 2004.
[10] H. C. Kimaro, “Decentralization and sustainability of ict based health
information systems in developing countries: A case study from tanza- nia,”
2006.
[11] R. B. Heeks, “Health information systems: Failure, success and local
improvisations,” International Journal of Medical Informatics, vol. 75, p.
125137, 2006.
[12] R. Hebert, “Health informatics - where to start? national e-health options for
developing countries,” Information Technology in Developing Countries,
IFIP WG 9.4, vol. 18, no. 1, 2008.
[13] A. Martinez, V. Villarroel, J. Seoane, and F. del Pozo”, “Rural telemedicine
for primary healthcare in developing countries,” IEEE transactions on
Information Technology in Biomedicine , Vol. 9, 66-72,, 2004.
[14] A. Rendon, A. Martinez, D. Lopez, and M. D. ”, “Rural telemedicine
infrastructure and services in the department of cauca, colombia,”
Telemedicine Journal and e-Health, Vol. 11, n. 4, 451-459, 2005.
[15] A. Martinez, V. Villarroel, J. Seoane, and F. del Pozo”, “An economic
analysis of the ehas telemedicine system in alto amazonas,” Journal of
Telemedicine and Telecare , Vol. 13(1), 7-14, 2007.
[16] I. Bebea González, “Diseño de un plan de sostenibilidad para redes de
comunicaciones rurales: estudio del caso Napo,” 2010.
[17] S. Sunden, “Information and communication technology applied for de-
veloping countries in a rural context. towards a framework for analysing
factors influencing sustainable use,” 2006.
[18] Batchelor, S. and Norrish, P., “Sustainable information and communication
technology (ict),” Gamos Ltd., 2003. [Online]. Available:
http://www.sustainableicts.org/sustainability1.htm
[19] S. Surana, “Beyond pilots: keeping rural wireless networks alive,” 5th
USENIX Symposium on Networked Systems Design and Implementation,
2008.
[20] S. Cisler, “Planning for sustainability: How to keep your ict proyect running
(schools online),” http://www2.ctcnet.org/ctc/Cisler/sustain.doc, 2002.
[21] “Centreon, network, systems and application monitoring based upon the
open source monitoring engine nagios,” http://www.centreon.com, 2009.
[22] E. Galstad, “Nagios 3.x documentation,” http://www.nagios.org, 2008.
[23] J. Vincent, RT Essentials. O’Reilly, 2005.
[24] A. Martinez, “Evaluación de impacto del uso de tecnologías apropiadas de
comunicación para el personal sanitario rural de paises en desarrollo,” 2003.